Background: The mean age of patients starting dialysis increased over the years, as has the proportion of patients with diabetes mellitus, ischaemic heart disease, peripheral vascular disease (PVD), cerebrovascular disease (CD) and malignancy. We assessed dialysis modality choice within subgroups of patients with these comorbidities and in different age categories and subsequently evaluated the association between modality choice and patient survival in these subgroups.
Methods: Seven European renal registries participating in the ERA-EDTA Registry provided data from 15,828 incident peritoneal dialysis (PD) and haemodialysis (HD) patients (1998-2006) with available comorbidity data. The likelihood to receive PD rather than HD was assessed with logistic regression and 3-year survival on PD versus HD was evaluated using Cox regression.
Results: Besides large international variations in the likelihood to receive PD, we found that elderly patients and patients with PVD, CD, malignancy and multiple comorbidities were significantly less likely to receive PD than HD. Overall patients starting on PD had survival benefits [adjusted hazard ratio (HR(adj)) 0.82 (0.75-0.90)], especially patients without comorbidity [HR(adj) 0.65 (0.53-0.80)] or those with malignancy [HR(adj) 0.73 (0.56-0.94)]. In males, survival benefits of PD were independent of diabetic status. Conversely, diabetic females tended to have increased mortality risk on PD [HR(adj) 1.16 (0.93-1.44)], especially if they were >70 years [HR(adj) 1.55 (1.15-2.08)].
Conclusions: In general, modality choice was consistent with expected survival. However, elderly patients, non-diabetic patients and those with malignancy were less likely to receive PD, even though they had decreased mortality risk on PD. Also, although a survival benefit of PD was found for male patients without comorbidity, HD was just as likely to be the chosen dialysis modality as was PD for these patients.